heroism franco blau zimbardo

8/3/2019 Heroism Franco Blau Zimbardo

1/16

Review of General Psychology

Heroism: A Conceptual Analysis and DifferentiationBetween Heroic Action and AltruismZeno E. Franco, Kathy Blau, and Philip G. Zimbardo

Online First Publication, April 11, 2011. doi: 10.1037/a0022672

CITATION

Franco, Z. E., Blau, K., & Zimbardo, P. G. (2011, April 11). Heroism: A Conceptual Analysis

and Differentiation Between Heroic Action and Altruism. Review of General Psychology.

Advance online publication. doi: 10.1037/a0022672


2/16

Heroism: A Conceptual Analysis and Differentiation Between HeroicAction and Altruism

Zeno E. Franco and Kathy BlauPalo Alto University

Philip G. ZimbardoPalo Alto University and Stanford University

Heroism represents the ideal of citizens transforming civic virtue into the highest form of civic action,

accepting either physical peril or social sacrifice. While implicit theories of heroism abound, surprisingly little

theoretical or empirical work has been done to better understand the phenomenon. Toward this goal, we

summarize our efforts to systematically develop a taxonomy of heroic subtypes as a starting point for theory

building. Next we explore three apparent paradoxes that surround heroismthe dueling impulses to elevate

and negate heroic actors; the contrast between the public ascription of heroic status versus the interior decision

to act heroically; and apparent similarities between altruism, bystander intervention and heroism that mask

important differences between these phenomena. We assert that these seeming contradictions point to an

unrecognized relationship between insufficient justification and the ascription of heroic status, providing more

explanatory power than risk-type alone. The results of an empirical study are briefly presented to provide

preliminary support to these arguments. Finally, several areas for future research and theoretical activity are

briefly considered. These include the possibility that extension neglect may play a central role in publics view

of nonprototypical heroes; a critique of the positive psychology view that heroism is always a virtuous,

prosocial activity; problems associated with retrospective study of heroes; the suggestion that injury or death

(particularly in social sacrifice heroes) serves to resolve dissonance in favor of the heroic actor; and a

consideration of how to foster heroic imagination.

Keywords: heroism, altruism, moral courage, banality of heroism, heroic imagination

Heroism is frequently viewed as an apex of human behavior;

watching a heroic act is compellingliterally commanding our

attention. We often feel that while we as individuals would like to

achieve heroic status, this goal must be a remote possibility re-

served for an elect few with special skills or luck. Heroism is a

concept that is simple at its surface. A straightforward definition

that isat firstsatisfying is to act in a prosocial manner despite

personal risk. However, this surface masks a number of subtle,

interrelated paradoxes that arguably make heroism one the most

complex human behaviors to study. Further, it seems likely that

the contradictory nature of heroism is precisely what makes it

compelling. Heroism is a social attribution, never a personal

one; yet the act itself is often a solitary, existential choice. It is

historically, culturally and situationally determined, thus heroes

of one era may prove to be villains in another time when

controverting evidence emerges; yet some heroes endure across

the centuries. Moreover, the very same act accorded hero status

in one group, such as suicide bombing, is absolutely abhorrent

to many others.

Throughout this work, we advance four primary ideas: (a)

The concept of heroism is a way to unify several types of

courageous or brave actions that have largely been treated

independently in the literature to date; (b) that the simple

presence of risk accompanying prosocial behavior is not enough

to define heroism; (c) heroism is viewed as distinct from other

prosocial activities, such as compassion and altruism (and may

represent an entirely different behavior); and (d) that while

heroism is primarily a positive and prosocial act, a simplistic

view of this behavior misses important (and sometimes nega-

tive) aspects of the phenomenon.

In an effort to present these ideas in a logical fashion, while also

allowing navigation room for exploration, we organized these

thoughts in three sections: First, we summarize our initial efforts atdistilling implicit views of heroism to set the stage for this analysis.

Second, a somewhat broader conceptual analysis of heroism is offered

to explore both the paradoxical nature of heroic action and to begin

the process of more rigorous theorizing in this arena. Third, using the

findings of a preliminary, Internet-based study designed to evaluate

our reflections on implicit theories of heroism held by the general

public, we attempt to situate the specific idea of heroism (rather than

related concepts) within the context of existing psychological theories.

Finally, the discussion offers several arenas for further inquiry that we

feel may serve to more deeply explicate the phenomenon of heroic

action and its social construction.

Zeno E. Franco and Kathy Blau, Pacific Graduate School of Psychol-

ogy, Palo Alto University; Philip G. Zimbardo, Pacific Graduate School

of Psychology, Palo Alto University, Emeritus Professor, Stanford

University.

Zeno E. Franco is now at the Department of Family & Community

Medicine, Medical College of Wisconsin.

We would like to acknowledge the generous financial support of the

Center for Compassion & Altruism Research & Education (CCARE) at

Stanford University, Dr. James R. Doty, Director, which made portions of

this work possible. Dr. George A. Quattrone contributed significantly to

revisions to this paper, sharpening both the theoretical and analytical

elements of this work. Matt Langdon, Director of the Hero Construction

Company, assisted with participant recruitment.

Correspondence concerning this article should be addressed to Zeno E.

Franco, Department of Family & Community Medicine, Medical College

of Wisconsin, 8701 Watertown Plank Road, Box 26509, Milwaukee, WI

53226. E-mail: [email protected]

Review of General Psychology 2011 American Psychological Association2011, Vol. , No. , 000000 1089-2680/11/$12.00 DOI: 10.1037/a0022672

1


3/16

Our Prior Work on Heroes

A brief review of our prior work in this area is in order. We have

proposed a largely situationist view of heroism, asserting that contrary

to the idea of the heroic elect (Hughes-Hallet, 2004), most people are

capable of heroism with the right mindset and under certain condi-

tions that call for heroic action. This idea, the Banality of Heroismargument initially offered as an essay response to the EDGE 2006

annual question, what is your dangerous idea? (Zimbardo, 2006)

and later expanded in other works (e.g., Franco & Zimbardo, 2006

2007; Franco & Zimbardo, 2006; Zimbardo, 2007b), recalls Hanna

Arendts (1963) Eichmann in Jerusalem: A Report on the Banality of

Evil, which underscored the latent ability of even seemingly normal

people to commit horrific acts. The banality of heroism argument

takes us in the opposite moral direction, but similarly asks the ques-

tion, what if the capability to act heroically is also fundamentally

ordinary and available to all of us? In turn, this position led us to

carefully reexamine the history of the ideal of heroic action, the

categories of heroic types that are widely recognized by the public,

and to explore the particular situations in which heroism seems mostlikely to occur.

The Heroic Core in Overlapping Forms of Courage

To date, psychologists have seemed almost reluctant to confront

heroism directly, instead discussing the phenomenon in terms of

civil courage, courageous resistance, extreme altruism, moral cour-

age, and the like. While some of these terms focus more on

physical heroes and others on what Eagly and Becker (2005)

termed cultural heroes, we assert that each of these concepts

point to a shared, core idea of heroic action (Greitemeyer, Oss-

wald, Fischer, & Frey, 2007; Shepela et al., 1999). In part, this

stance reflects the criticisms offered by Martens (2005), whosuggested that Becker and Eagly (2004) had examined prototypical

physical risk heroes at the expense of addressing other forms of

principle-driven courage that could equally be termed heroism.

In fact, historical views of heroism underscore the importance of

nobility of purpose or the principle underlying the heroic act (see

Zimbardo, 2007b). This sentiment is echoed in dictionary entries

from the early part of the last century. For instance, Websters,

1913 dictionary stated, Heroism . . . is a contempt of danger, not

from ignorance or inconsiderate levity, but from a noble devotion

to some great cause, and a just confidence of being able to meet

danger in the spirit of such a cause (emphasis added; Olson &

LaRowe, n.d. from definition footnotes, pp. 334 & 689). While

there are obvious differences between the various forms of braveryand courage, there are also several elements that can be viewed as

binding them together conceptually. First, each involves a level of

peril or sacrifice that goes well beyond what is expected in other

prosocial behaviors. Second, each entails a willingness to enter a

fraught situation despite clear barriers to entry and obvious paths

of exit. Third, across all forms of heroics, the actor must transcend

considerable fear to act decisively (Franco & Zimbardo, 2006

2007; Zimbardo, 2007b). From this position we attempted to

reformulate the types of heroism more comprehensively, describ-

ing three broad forms of heroic action: martial (military) heroism,

civil heroism, and social heroism (Zimbardo, 2007b).

Martial Heroes: Archetypal Antecedents and Modern

Military Heroism

Even contemporary research in this area has used terms like cour-

age and heroism synonymously (Rate, Clark, Lindsay, & Sternberg,

2007). While the meanings are very close, older dictionaries were at

pains to decompose these ideas (See Olson & LaRowe, n.d., 1913

Websters Revised Unabridged Dictionary). Courage was seen as

necessary but insufficient to meet this archaic standard of heroism. In

military terms, for example, a soldier might courageously face death,

but this differed from choosing to go into the heart of battle or to act

above and beyond the call of duty (for a detailed discussion, see

Zimbardo, 2007b, pp. 461462). This willingness to take conspicu-

ous, bold action in a way that sets one apart from his already brave

peers continues to serve as the high-water mark of heroism in modern

warfare1 (Armed Forces Act, 2000; Glanfield, 2005; Murphy, 2005;

United States Army, 2005). The ideal of the war hero is clearly

echoed in other contexts, including those who more routinely risk life

and limb in the line of duty and who are bound to a code of conduct,

such as police officers, firefighters and paramedics (Zimbardo,

2007b). Collectively, we also refer to this group as physical risk,duty-bound heroes.

Heroism in the Civilian Sphere

Civil heroism is similar to martial heroism because it involves

physical peril. However, there is no military code of conduct to fall

back upon, the actor may not be trained to deal with the situation,

and there is no specific script (e.g., an honor code) that guides the

individual toward heroic actionas is the case for martial heroism

(Zimbardo, 2007b). Thus, the standard for duty-bound and non-

duty bound physical-risk heroism differs, but the style of engage-

ment and potential sacrifices are comparable. Death, serious injury,

disfigurement, and pain are all possible outcomes of acting on

behalf of others in jeopardy. (Other authors have referred to thisaction as civil courage, but have not consistently distinguished

between physical peril and social sacrifice e.g., Greitemeyer et al.,

2007). A classic example is a civilian bystander who performs an

emergency rescue. We also refer to this form of heroism as

physical risk, non-duty-bound heroism later in this discussion.

Social heroism, in contrast, typically does not involve immedi-

ate physical peril. It is nonetheless associated with considerable risk

and personal sacrifice in other dimensions of life, including serious

financial consequences, loss of social status, possible long-term health

problems, and social ostracism (Glazer & Glazer, 1999; Shepela et al.,

1999). The goal of social heroism can be seen as the preservation of

a community-sanctioned value or standard that is perceived to be

under threat. In some cases, the actor is actually trying to establish aset of extra-community standardspushing toward a new ideal that

has not yet found wide acceptance. This is an important distinction, as

hindsight bias makes it easy to assume that the value asserted by the

heroic actor has always been an accepted social standardwhich is

often not the case. What we call social heroism here, to underline its

similarities to martial and civil heroism, has also been conceived of by

others as courageous resistance (Shepela et al., 1999), rescue

1 Note that only 464 U.S. combatants in World War II received the

Medal of Honor, and of these just 211 had survived the engagement for

which the Award was conferred (Murphy, 2005).

2 FRANCO, BLAU, AND ZIMBARDO


4/16

altruism (Oliner & Oliner, 1988), moral rebels (Monin, Sawyer, &

Marquez, 2008), or moral courage (Staub, 2011).

Further, although physical peril is not the most prominent fea-

ture of this form of heroism, involvement in prolonged social

heroics may lead to eventual or insidious physical risk (Glazer &

Glazer, 1999). By contrast, physical risk (martial and civil) heroes

are most often viewed as the archetypal heroic figures (Eagly &Becker, 2005; Hughes-Hallet, 2004). Their actions are usually

dramatic, occur rapidly, and are comparatively free from contro-

versy. Physical risk heroism usually involves some probability of

serious injury or death, but not certainty that either will occurin

fact the actors calculus may involve the expectation of a reason-

able chance of exiting the situation unscathed. With some excep-

tions, the individual performing the act is usually completely

removed from peril after a short period of time if they survive.

Social heroism is typically less dramatic, unfolds over a much

longer time period, and is frequently undertaken in private rather

than public settings. Yet, the costs associated with social heroism

are often certain to occur and are willingly engaged in over these

extended time frames (Franco & Zimbardo, 20062007; Glazer &

Glazer, 1999). From this perspective, it has been argued that socialheroism should be viewed as more heroic than physical risk forms

of heroism (Howerth, 1935; Peterson & Seligman, 2004).

A Preliminary Attempt at an Operational Definition

of Heroism

The assertion that these forms of courage share a heroic core,

modifies Becker and Eaglys (2004) characterization in two ways.

First, it suggests that the recognized risk and anticipated sacrifice

should not be limited to immediate threats to physical integrity or

death. Instead, it should be more comprehensive, also considering the

sacrifices typically encountered in socially courageous actions. Sec-

ond, because this broadens the idea of heroism to allow a multidi-

mensional view of the phenomenon, it also seems necessary to rule

out some forms of apparent heroism that might not, in fact, be heroic.

Thus, anticipated gain at the time of the act necessarily disqualifies it

from being heroic. However, if gains are subsequently accrued with-

out prior anticipation or motivation to attain them, the act should still

be upheld as heroic (Glazer & Glazer, 1999).

Our definition of heroism is as a social activity: (a) in service to

others in needbe it a person, group, or community, or in defense

of socially sanctioned ideals, or new social standard; (b) engaged

in voluntarily (even in military contexts, heroism remains an act

that goes beyond actions required by military duty2); (c) with

recognition of possible risks/costs, (i.e., not entered into blindly or

blithely, recalling the 1913 Websters definition that stated, not

from ignorance or inconsiderate levity); (d) in which the actor iswilling to accept anticipated sacrifice, and (e) without external gain

anticipated at the time of the act.

The Relationship Between 12 Heroic Types and

Situations That Call for Heroism

Using this conceptual definition of heroism and an elementary

model of heroic action (Zimbardo, 2007b, p. 480 & 482), we searched

for figures who exemplified this broad conceptualization of heroism

and who were deemed to be heroic by the general public and the

media drawing from current and historical newspaper sources, books,

and TV coverage focusing primarily on the last 100 years, and

occasionally on older examples where instructive. Next, we attempted

to distill this list into a manageable set of categories of heroic action.

Finally, we felt that while a number of dispositional attributes might

influence ones willingness or inclination to act in a heroic manner

(e.g., Walker, Frimer, & Dunlop, 2010), the situational factors in-

volved in heroism had not received sufficient attention and we at-

tempted to draw a connection between the proposed heroic types andthe situations that drove their heroic action (Franco & Zimbardo,

2006; Glazer & Glazer, 1999; Shepela et al., 1999; Zimbardo, 2007b).

This effort resulted in a preliminary taxonomy of 12 heroic subtypes

and the situations that call for heroic action (for a detailed expla-

nation of the taxonomy, including definitions for each condition, risks

and sacrifices encountered, and individual exemplars, see Zimbardo,

2007b, pp. 468471). Our taxonomy, summarized in Table 1, in-

cludes martial and civil heroism (the two forms of heroism involving

physical risk), as well as 10 variations of social heroism and their

initiating, defining situations.

As is the case with most a priori and informal taxonometric

processes, the resulting list of heroic types and situations is some-

what arbitrary, reflecting the considered views of the authors. It

should be noted in the process of developing this taxonomy, therelative merits of various heroic exemplars were debated at

lengthrevealing the importance of factors such as intention,

success versus failure, presence or absence of onlookers, and so

on. This served as a constant reminder of the degree to which the

title hero is a social construction (Rankin & Eagly, 2008) that

may or may not accurately reflect the actual merits of an individ-

uals actions. Though informal, we view this process as a form of

disciplined imagination (Wieck, 1989) intended to stimulate

wider exploration and critical discussion of the nature of heroism.

As such, our working taxonomy is necessarily open to further

development, refinement, and modification as part of the theory

building process. Interestingly, as our work has progressed, we

have encountered prior taxonomies that suggest these heroic cat-egories are relatively stable over time and across researchers.3

Moving From Implicit Theories of Courage Toward

an Explicit Theory of Heroism: A Conceptual Analysis

In the prior section we presented an attempt to bring greater

semantic clarity to the study of heroism by specifying the types of

situations in which the term is used both by laity and researchers

to illuminate the breadth of the conceptual space occupied by

heroism, and by defining our terms. This is one of several ap-

proaches used in conceptual analysis (Machado & Silva, 2007).

The presentation of three apparent paradoxes (using the term

2 For example, a soldier volunteering for an especially dangerous mis-

sion, where all are told not volunteering is completely acceptable.3 For example, drawing on his own dissertation, Orrin Klapp (1954)

stated, The most common patterns or roles of heroism include: the

Conquering Hero, an invincible figure of power who seemingly can do

anything; the Clever Hero, a trickster who wins by cunning rather than

force; the Cinderella or Unpromising Hero; the Quest Hero or idealistic

seeker; the Deliverer, Defender or Avenger; the Popular Benefactor; the

Culture Hero; and the Martyr (p. 58). Also see Klapp (1948) for a more

extensive discussion of the topic. More recently, just as this article was

being submitted for publication, Allison and Goethals (2010) published a

book on heroes that reflects similar categories.

3HEROISM: A CONCEPTUAL ANALYSIS


5/16

loosely) allows us to extend the conceptualization by relying on

several of the strategies advanced by Machado and Silva as a part

of the theorizing process. In particular, early efforts to characterize

heroism (including our own) suffer from the nominal fallacy

(calling something heroism does not explain the phenomenon),

a lack of acknowledgment of subtle steps in the ascription of

heroic status (which we begin to elucidate further in the first and

second paradoxes), and unjustified extension of familiar concepts

to an unfamiliar domain (addressed in detail in the third paradox).

On their own, each of these observations provide some insight, but

we also bring them together to suggest an explanatory model thatgoes beyond the simple altruism plus risk argument.

Paradox 1: Elevation and Negationthe Role of

Dissonance

When we see an individual acting with courage despite the

risksfor example, the televised images of an average person

racing into a burning building to rescue a child, we get a lump in

our throat, are filled with a sense of the heights of compassion that

humans can reach, and we might even wonder if we could aspire

to such action if given that same set of circumstancesa sensation

akin to elevation (Algoe & Haidt, 2009). However, this is after the

fact and at distance, with the successful outcomes known. An

observer at the scene might very well describe the person running

into that dangerous situation as foolhardy (Klapp, 1954).

Similarly, a news story about a corporate whistle-blower may

lead the viewing audience to feel affinity and to acclaim the heroic

figure, while those actually working with her experience disgust at

her unwillingness to conform to established norms within the

organization and keep the situation under internal wraps. The

actions of the moral rebel can be conceptualized as offering

implicit reproach toward her coworkers. In turn, the rebel isrejected for not being in communion with others in the organiza-

tionan almost preemptive move on the part of observers that

stabilizes feelings of moral adequacy despite inaction (Monin et

al., 2008). Monin and colleagues (2008) point to a fairly dichoto-

mous outcome in which uninvolved observers ascribe higher rat-

ings of morality and agency to moral rebels, while vested actors

castigate these same individuals.

However, the picture is probably more nuanced. Anecdotal

evidence, classical literature, and prior theorists all suggest that

reactions to heroism in both the unvested public and vested non-

heroic actors are fickle (Klapp, 1954; Siklova, 2004). Even un-

Table 1

Twelve Heroic Subtypes and Situations That Call Forth Heroic Action

Risk type Heroic subtype Definition/situation

Physical Peril 1. Military and other duty-bound physical risk heroes Individuals involved in military or emergency responsecareers that involve repeated exposure to high-risksituations. Heroic acts must exceed the call of duty.

2. Civil heroesnonduty bound physical risk heroes Civilians who attempt to save others from physical harm ordeath while knowingly putting their own lives at risk.

Social Sacrifice 3. Religious figures Dedicated, life-long religious service embodying highestprinciples or breaking new religious/spiritual ground.Often serves as a teacher or public exemplar of service.

4. Politico-religious figures Religious leaders who have turned to politics to affectwider change, or politicians who have a deep spiritualbelief system that informs political practice.

5. Martyrs Religious or political figures who knowingly (sometimesdeliberately) put their lives in jeopardy in the service of acause or to gain attention to injustice.

6. Political or military leaders Typically lead a nation or group during a time of difficulty,such as a war or disaster. Serve to unify nation, provideshared vision, and may embody qualities that are seen asnecessary for the groups survival.

7. Adventurer/explorer/discoverer Individuals who explore unknown geographical areas or

use novel and unproven transportation methods.8. Scientific (discovery) heroes Individuals who explore unknown areas of science, use

novel and unproven research methods, or discover newscientific information seen as valuable to humanity.

9. Good Samaritan Individuals who are first to step in to help others in need.Situation involves considerable disincentives for altruism.May/may not involve immediate physical risk.

10. Odds beater/underdog Individuals who overcame handicap or adverse conditionsand succeed in spite of such negative circumstances,thereby provide a social, moral model for others.

11. Bureaucracy heroes Employees in large organizations in controversialarguments within or between agencies. Typically,involves standing firm on principle despite intensepressures to conform or blindly obey higher authorities.

12. Whistleblowers Individuals who are aware of illegal or unethical activitiesin an organization who report the activity publicly to

effect change, without expectation of reward.



6/16

vested observers may be quick to negate the heros acts at the

slightest hint of countervailing information about the heros integ-

rity, motives, or intentionseven if these aspersions have no real

bearing on the heroic act itself. There is a constant tension between

the desire to elevate and the desire to castigate the actions of

heroesespecially social heroes because their actions are easily

viewed as threatening, but also with physical risk heroes who havea checkered history (Franco & Zimbardo, 20062007).

Paradox 2: The Public Stage Versus the Interiority of

the Heroic Decision

Our reliance on social construction to identify whom among us

is a captain among men points out another disconnection in our

understanding of what is heroic. It can be argued that the action or

behavior ultimately stands as heroic or not in the absence of any

social milieu (Howerth, 1935). In fact, some of the most heroic

actions in history are the work of the so-called unsung hero,

those whose acts never come into the limelight. Military reviews ofheroism acknowledge that the most heroic martial actions are

undertaken in circumstances that do not favor overt recognition;

situations in which no survivors are left to recount the action that

took place, events that are forgotten because the required forms are

not completed by commanding officers (a written record of the

action is generally needed for later conferral of military honors), or

cases in which soldiers with better political connections are ac-

knowledged instead of those closest to the heart of combat (Mur-

phy, 2005; Quaife, 1931). It should also be evident that heroes are

more likely to be found among conquerors than those conquered,

and, with a few notable exceptions, by those who are literate and

who maintain written records for historical consumption than from

those with oral traditions, and so on.This points to an underlying premise we think is important to

acknowledge: While heroism is generally considered to be a proso-

cial behavior and the act must be witnessed or evaluated by

spectators to receive acclaim, the decision to act in a heroic

manner does not necessarily emanate from prosocial motivation,

nor does it require an audience. In fact, the decision to act is almost

always a private, interior process that occurs before and in the

absence of public knowledge about what is about to unfold. For

many heroes, engaging heroically may have more to do with the

individuals examination of their own deeply held standards for

behavior in a given situation (Shepela et al., 1999), and that despite

profound reservations about the actions they are about to take,

these internalized standards are considered to be so inviolable as tocompel action even in the face of peril (Glazer & Glazer, 1999;

Shepela et al., 1999; Tangney & Dearing, 2002).

The intensely public nature of the ascription of heroic status

perhaps viewed by millions as a viral video on the Internetbelies

the profound interiority of the decision to act in a heroic manner.

While the hero may later be lauded by throngs, this moment of

decision is often taken in complete aloneness, even if it is in the

presence of others. Moreover, we argue that this tension between

the publics interpretation of the event and the private decision

process of the hero holds at least part of the key to understanding

the social ascription of heroic status.

Paradox 3: Altruism Versus HeroismDifferences in

Risk, Velocity, and Barriers to Entry

The assertion that heroism is simply an extension of prosocial

behaviors, such as altruism, is initially compelling and probably

captures some of the contours of the phenomenon. However, it is

worth noting several limitations of this perspective (see, e.g.,Greitemeyer et al., 2007). Shepela and her colleagues underscored

the importance of recognizing the limitations of altruism in ex-

plaining heroic behavior, arguing that, A theory of courageous

resistance must account for the observation that not all altruistic

individuals were willing to engage in courageous resistance, and

that even those who do so do not resist at every opportunity

(Shepela et al., 1999, p. 799). To further illuminate this observa-

tion, we offer several conceptual distinctions between altruism and

heroism.

First, we assert that the level of risk incurred in altruism is

considerably lower than the minimum risk threshold for heroic

status. A few theorists have added a self-sacrifice criterion to the

definition of altruism, bringing it closer to the notion of heroism,

maintaining that altruism is defined by the helper incurring somecost for their helping act. This conception of altruism plus modest

risk is anecdotally understood as a situation in which everyone

should act, but only some people do (paraphrasing a sentiment

expressed by Wesley Autrey as he demurred on the title of hero

after dramatically rescuing a disabled man who had fallen onto the

tracks New York subway; Cardwell, 2007).

Second, while prosocial behaviors such as volunteering appear

to protect health in part at least by reducing social alienation (e.g.,

Oman, Thoresen, & McHahon, 1999; Thoits & Hewitt, 2001),

social heroism often leads to the opposite outcomerejection by

valued social networks (Oliner & Oliner, 1988; Shepela et al.,

1999). Similarly, while physical risk heroes may experience some

later benefits, anecdotal accounts of increased suicide risk amongheroic rescuers after their celebrity has faded are common

(Braudy, 1997; Hopkins & Jones, 2003). The outcomes for the

heroic actor are quite different than altruistic actors.

Third, most bystander research has been undertaken in the absence

of real risks and rarely systematically manipulates the perception of

risk to the participant, instead typically focusing the perceived threat

on a confederate. Indeed, Shepela et al. (1999) noted, It is difficult,

indeed, very likely unethical, to test courageous resistance empir-

ically . . . analog studies would not tap the motivations which

support the leap to which courageous resistance entails (p. 801).

We are aware of only one study (Fischer, Greitemeyer, Pollozek,

& Frey, 2006) that systematically varies the level of emergency-

related danger focused on the participant, which addresses the

risks of participant intervention in a way that is well aligned with

the conceptualization of heroism offered here. Interestingly, Fis-

chers study found that empathy and feelings of personal respon-

sibility were more central to general help-giving decisions (altru-

ism) than those involving civil courage. Fischer points to this

finding as an illustration of the danger in extrapolating from the

classical Latane and Darley (1970) helping model in an attempt to

explain courage.

Fourth, altruism and bystander intervention are typically accom-

panied by a period of deliberative indecision lasting from several

seconds in fairly straightforward situations to several minutes in

comparatively ambiguous contexts (Latane & Nida, 1981). In



7/16

contrast, anecdotal accounts of physical risk heroism suggest that

individuals are impelled to engaged in split-second decisions that

propel action despite situational complexity (Howerth, 1935; She-

pela et al., 1999). Recalling the Websters, 1913 view of heroism,

which touches on the impetuous or impulsive quality of heroic

action, we are reminded that martial and civil heroism may be

partially defined by successful execution within a very smallwindow of opportunity, and the ability to react in the moment is

one of the defining characteristics of heroism. For example, Klapp

(1954) observed, the same act performed too soon or too

late . . . may make a person a fool rather than a hero (p. 59).

Finally, where possible most bystanders will resolve conflicting

avoidance impulses by finding ways to avoid having to choose a

course of action (Latane & Nida, 1981, p. 309)bystander inter-

vention is much more likely to occur when psychological exit of

the situation is not easy. In contrast, heroic action is often under-

taken in the presence of clear paths to exitthe situation and despite

the factors that are typically associated with the diffusion of

responsibility (large crowds, passive bystanders, etc.; Fischer et

al., 2006). Again, it is instructive here to refer to the definition ofgallantry, which suggests that heroes are willing to step into the

fray and deliberately approach dangerous situations despite the

fact that barriers to entering these situations are steep (Shepela et

al., 1999).

Heroic Status: The Role of Choice and Insufficient

Justification

Taken together, these apparent paradoxes begin to point, at

minimum, to an extension of the altruism plus risk argument that

we believe is more nuanced and offers greater explanatory power.

Witnessing a heroic actin which an individual enters a danger-

ous situation despite clear avenues for exit and substantial barriersto entrymust leave the observer with mixed feelings, indeed the

action may seem irrationally risky (Walton, 1986). Further, be-

cause the actors choice is deeply interior, the rationale is not

readily accessible to the observer, leaving the impression that the

action was not sufficiently justified (Festinger & Carlsmith, 1959).

Yet, the observer must still make an attribution about the action

thus opening the situation to a wide range of interpretations

(Klapp, 1954). If the actor is unsuccessful, he is quickly deemed a

fool for having risked so much. However, in the very same

situation in which chance looks favorably upon the actor, she is

lauded as a hero for having succeeded despite the apparently

unacceptable level of risk taken. Either interpretation successfully

resolves the dissonance the situation has produced in observers.From this vantage, we can assert that risk involved in heroism is

one that the actor was under no moral obligation to acceptone is

not morally or ethically bound to act in ways that will place ones

self in imminent danger (Shepela et al., 1999). The role of insuf-

ficient justification in the ascription of heroic status is hinted at by

Ludvik Vaculik, a Czech dissident reflecting on the resistance

process, who stated, Heroic deeds are alien to everyday life, they

flourish in exceptional situations, but these must not be of long

duration (Vaculik, 1990, p. 32). Thus, we can say in contrast to

altruism, heroism is a situation in which, no one is should act, but

a few do anyway.

Summary of Key Points and Hypotheses

First, where others (Eagly & Becker, 2005; Shepela et al., 1999)

have suggested a fundamental distinction between prototypical

physical risk heroism found in martial and civil heroes and prin-

ciple driven social heroism, we argue that there are fundamental

elements of all three forms of action that are similar (e.g., accep-tance of various forms of risk, willingness to step into the fray,

etc.) and that all three forms of action point to a single, underlying

heroic ideal held in common. To begin addressing this question,

we tested how the general public would categorize the 12 heroic

types we originally proposed.

Second, we speculated that given the duration and certainty of

risk associated with social heroism, individuals engaging in this

type of activity would be viewed as at least as heroic as the more

readily accessible forms of martial and civil heroism. Further, we

left open the possibility that in a few cases social heroes might be

viewed as more heroic than the prototypical versions of heroism

involving immediate physical risk (Howerth, 1935; Martens, 2005;

Peterson & Seligman, 2004).

Third, we postulated that the ascription of heroic status on thepart of uninvolved observers revolves not simply around the pres-

ence of any risk while performing a prosocial act, but the disso-

nance caused by the actors choice to take on what is seen as an

irrational or unjustified level of risk.

Fourth, we asserted that heroism is at minimum a special case

within prosocial behaviors. At maximum, we speculated that her-

oism encompasses a fundamentally different class of behaviors

that, while sharing some overlap with altruism, represents a dif-

ference in kind rather than a difference in degree (Greitemeyer et

al., 2007). This is in contrast to Shepela (1999) and others who

have posited that heroic behaviors are a natural subset of altruism.

Method

To evaluate current public perceptions of heroism, to examine

potential differences between heroism and altruism, and to begin a

more systematic evaluation of the 12 heroic subtypes offered here,

a pilot survey was developed. Because of the number of items we

were interested in, the survey was divided into two forms, A and

B to ensure that participants would not be overburdened with

survey length.

The first portion of the survey was constructed by writing brief

hypothetical scenarios based on the 12 proposed heroic subcate-

gories, which we refer to as generic situation items. These

prompts ostensibly contained the essence of the heroic action

without the potentially confounding factors of whether or not

participants recognized particular heroes, or the emotional valence

of particular stories that might distract from the core behaviors.

The two physical-risk categories, martial (duty bound) heroism

and civil (nonduty bound) heroism were further broken down into

five martial and five civil heroism items to provide a deeper

understanding of what is viewed as central to the heroic construct

within each of these domains. Ten social heroism items were taken

directly from the taxonomy. Participants responded to each situa-

tion item with a forced choice between three alternatives: heroic,

altruistic, or neither heroic, nor altruistic.

Next, a matching set of prompts was developed by identifying

an actual figure who had been publicly conferred some level of



8/16

heroic status. The individual was named and a brief description of

the situation and actions taken was provided. Each of these nar-

rative items was drawn from exemplars offered in the initial

taxonomy (Franco & Zimbardo, 2006; Zimbardo, 2007b), histor-

ical accounts, Medal of Honor recipients, Carnegie Medal Award-

ees, and news stories. While the items were designed to address the

same context described in the situation items, some categories hadvariations in wording across the two item types. For example, a

generic situation might offer, A civilian rescues someone from a

burning building, while the matching narrative item stated, Rob-

ert G. Falconer rescued Fred Johnson from burning in his home.

Seeing flames in Johnsons house from his home across the street,

Falconer went into Johnsons home and dragged Johnson out of

the house.

To develop a finer grained analysis of how our participants

viewed each of the items, the generic situation items and narrative

items were presented as Likert-type questions. Participants re-

sponded to a 5-point scale, anchored by not at all heroic and

extremely heroic at opposite poles with somewhat heroic as

the midpoint.A final question, which appeared on all forms of the survey,

stated simply Do you feel there is a difference between altruism

and heroism? followed by two open-ended items, asking, If, so

what is the difference? and If you had to define heroism, what

would your definition be? These items were designed to tap into

the participants implicit views of altruism and heroism.

Though the intent of this study is illustrative rather than de-

signed to establish a validated measure of heroism, a number of

items were included to show at least a reasonable level of face

validity and discriminant validity for the heroism items. Eleven

items were included that were designed to be marked altruistic to

the exclusion of the alternatives, and 19 items were intended to

prompt a neither heroic nor altruistic response. All altruisticitems were predominantly marked in the expected direction. Six-

teen of the 19 neither items were also marked in the expected

direction, with the remaining three items being viewed as more

altruistic than anticipated.4

Procedures

Participants were recruited using convenience sampling by ad-

vertising the survey on several Websites. At first the Websites used

were related to the topic of heroism and morality.5 Finally, in an

effort to obtain a more varied and representative sample of partic-

ipants not already interested in topics associated with heroism, aconcerted effort was made to advertise the survey on a popular

blog that has high traffic and international reach. These efforts led

to the survey being advertised on a popular technology blog.6 The

gross majority of the participants were directed to the survey from

the technology blog. Participants who navigated to the survey

Website were randomly assigned to either form A or form B, and

next assigned to one of four question order presentations. This

allowed for a Latin Square presentation of questions within each

form to control for ordering effects. Completion of the survey took

30 min. Participants had the option of exiting the survey at any

time. No compensation was offered.

Participants

A total of 3,696 adults aged 18 and over responded. Because this

was an Internet study, some participants viewed the consent form

(thus creating an observation), but failed to complete any of the

survey items. Individuals who did not complete the initial 5 ques-

tions were removed from analysis to exclude these incompleteobservations. Visual inspection of the data showed that this was a

reasonable decision rule that removed nearly all nonresponses,

while including others who responded to at least some of the items.

Participant demographics and comparisons between those who

were assigned to Form A and Form B are summarized in Table 2.

Analyses

Given that there were no significant differences between the

demographic characteristics across Form A and B, the data were

combined for analysis and display. A series of chi-square goodness

of fit tests were conducted examining each of the situation items to

measure whether the item was predominantly viewed as heroic,altruistic, or neither. Hypotheses 2 and 3 were also addressed

through a regression model. Hypothesis 4 was addressed using the

final survey item, which asked participants to state if they felt there

was difference between heroism and altruism using chi-square and

representative quotes pointing to potential themes in open-ended

responses.

Results

Hypothesis 1: Core of Heroic Action Extends

Beyond Prototypical Physical Risk Heroism to

Encompass 12 Heroic Subtypes

Univariate analyses suggest that archetypical situations in the

martial, civil, and social heroism domains are all seen as predom-

inantly heroic. Interestingly, civilian fire rescue was viewed as the

most purely heroic, with 96% of participants ascribing the her-

oism category to this type of activity, while just 4% saw it as

altruistic, and less than 1% saw it as fitting in neither category. In

contrast, one of the most archetypal forms of military heroisma

solider laying down his life to allow others to escape, was seen by

88% of respondents as heroic, while about 9% saw this behavior as

altruistic, and 3% said it was neither.

The contrast between highest heroism responses for civil and

military heroism suggests that despite the strong historic associa-

tion between heroism and military service, current perceptions

ascribe more heroic value to actions that are taken when nospecific duty to serve exists. Two items involving criminality and

vigilantism were included to address the assertion offered here that

heroic actions taken should be more important than social attribu-

tions about the individuals intent or context. These included a

4 The items included to demonstrate discriminant validity for the hero-

ism concept are not discussed further here, but warrant future consideration

in the development of a validated heroism survey instrument.5 http://www.everydayheroism.org; http://heroworkshop.wordpress

.com, and http://greatergood.berkeley.edu/6 http://www.boingboing.net



9/16

criminal who risks his own life to protect others and a rogue who

ostensibly provides a prosocial service by capturing criminals.

While the benevolent criminal was viewed as a heroic actor, the

vigilante clearly was not. This suggests that in some situations

individuals with negative personal histories can be positively re-

appraised as heroic figures (Franco & Zimbardo, 2006 2007),

while in others the contextual information overrides the heroic

aspects of current actions. The data for civil heroism and martial

heroism situations are displayed in Figure 1.

Overall, however, the picture for social heroism is more mixed.

Even for those situations in which social activities are viewed as

heroic, they are generally seen as less heroic than civil or martial

heroism. Further, even for the most heroic situations in the social

arena, these activities are seen as less purely heroic and are

generally viewed as more altruistic. For example, in the politico-

religious figure item most closely associated with heroism (a

generic description of actions patterned after Dr. Martin Luther

King, Jr.), 74% stated this was an example of heroism, 18%

described it as altruism, and 7% felt the situation represented

neither category. Interestingly, the whistleblower situation was

viewed as heroic, but it was also clearly an item that resulted

mixed reactions. Twenty-six percent of respondents stated that

they viewed this form of action as neither heroic nor altruistic

perhaps reflecting the level of controversy often associated with

such individuals.

Finally, while we identified 10 forms of social heroism only 4 of

these situations were supported as predominately heroic by the

study participants. These included politico-religious figures, good

Samaritans, individuals fighting an unjust bureaucracy, and

whistleblowers. Six categories did not meet the criteria of beingpredominantly viewed as heroic. These included three subtypes

seen as largely altruisticreligious figures, scientific or discovery

figures, and odds beaters; and three categories viewed as neither

heroic nor altruisticmartyrs, political/military leaders, and ad-

venturers. The results of the situation items for social heroism are

summarized in Figure 2.

Hypothesis 2: Social Sacrifice Just as Heroic as

Physical Peril

In contrast to our assertions, the descriptive results offer clear

support for the notion advanced by Becker and Eagly (2004) that

physical risk heroes are viewed as most prototypically heroic.Social courageousness items were marked as heroic less fre-

quently, showed greater overlap with altruism, and were more

frequently viewed as being motivated by neither heroic nor altru-

istic intentions on forced categorization items. Further inferential

evidence rejecting the notion that social heroes are viewed as just

as heroic as physical risk heroes is offered in the next section.

These findings emphasize the role of physical risk as a determining

element in the publics view of prototypical heroic action.

Hypothesis 3: Insufficient Justification as an

Explanatory Factor in the Publics Ascription of

Heroic Status

As we considered this problem, it occurred to us that the 12

heroic categories we initially offered could be boiled down into

four categories that adequately described all forms of heroism

while also offering deeper insight into why some forms of action

are accorded greater heroic stature by the general public (Table 3).

The first two categories remained the same, martial (duty bound)

heroism and civil (non-duty-bound) heroism, both involving phys-

ical risk but under somewhat different conditions. Social heroism

was reconceptualized as having two primary forms (instead of the

initial 10 offered)those who defy systems (e.g., a bureaucracy

hero) and those who defy reality, that is, the limits of the known

world (e.g., an explorer who makes a discovery that changes

humanitys understanding of itself and/or the universe). Further, it

can be argued that there are similarities between those who defy

reality and martial heroics. Individuals in both categories have

elected to take this as a career path, have probably received some

training, and are following an established script that guides behav-

ior. In contrast, those who defy systems are more similar to

nonduty-bound physical risk heroes as they typically are not pre-

pared for this type of engagement before a presenting problem that

forces a decision about whether or not to take action and often

there is no established path that guides behavior toward heroism.

Thus, creating this distinction within social heroism provided an

avenue for us to explore the role of justifiability of risk across both

risk types (physical peril and social sacrifice).

Table 2

Participant Characteristics

Survey form

Characteristics A (n 1359) B (n 1364) p-valuea

Age

1821 123 1222229 368 3573039 400 395

4049 168 202 .50115059 75 8660 28 24Decline to state 197 178

Sex

.3584Female 477 481Male 690 715

Decline to state 192 168Ethnicity

White 959 1006Hispanic/Latino/Latina 27 30Asian American 32 23African American 7 9 .2794

Pac. Island/Nat. American 12 8Other 46 29

Mixed race 40 43Decline to state 236 216

Country or continentresponding fromUnited States 867 865Europe 134 140

Other North America 77 91Australia/New Zealand 55 58Asia 16 17 .9501South America 6 6Africa 6 5Decline to state 198 182

a For 2 test of homogeneity.



10/16

In an effort to test this hypotheses, the response matrix was

transposed so that we could examine potential predictors of the

averaged responses to the combined generic situation and narrative

ratings. Four items exploring the relationship between criminality

and heroism were included in the survey on exploratory basis (two

generic situation and two narrative items). As these were notconceptually representative of the prototypical physical risk hero,

nor described in the original taxonomy, these items were omitted

from the regression analysis. This allowed us to test the proposed

model that the risk type combined with perceived justifiablility

would predict ascribed heroic status better than risk type alone.

This approach also had the advantage of addressing the uninten-

tionally overpowered sample7, dramatically reducing the likeli-

hood of Type-I error, by attempting to predict the means for just 36

variables rather than scores across some 3,000 individuals, effec-

tively leaving us with only the most robust findings.

The data were organized by risk type (physical peril vs. social

sacrifice), and conditions that were assumed to violate levels of

acceptable risk (nonduty bound physical risk actors and those who

defy social systems) were coded into one group, while conditions

assumed to represent levels of risk that are more readily under-

stood as reasonable (duty-bound physical risk and those that at-

tempt to defy reality) were coded into a second group. The global

test of the model was significant, F(2, 33) 20.22, p .0001, and

physical peril significantly predicted mean ratings of heroism.

Moreover, as predicted, the covariate for justified risk was signif-

icant, with those in the justified risk category being rated as less

heroic as compared to those in the unjustified risk condition (see

Table 4). This model explained about 10% more of the variance

than one involving risk type as a single predictor of ascription of

heroic status.

Hypothesis 4: Heroism is Viewed as Different Than

Altruism

Again, referring back to Figures 1 and 2, clear distinctions

between heroism and altruism appear to be in evidence. For

example, almost all respondents see a civilian fire rescue as heroicand not altruistic. Perhaps most importantly, the Bystander breaks

up fight situation, which can be viewed as closely approximating

the extreme danger condition in the study conducted by Fischer

and his colleagues (2006) is viewed as heroic by only 46.5% of the

participants in the present study. This item is separated by al-

most 50 percentage points from the most acute form of civil

heroism, Fire rescuethus underscoring our view that current

experimental work on altruism in the face of substantial risk does

not fully capture the phenomenon of heroic action and suggesting

that further work is needed to explore these differences.

Further, the final survey question, which appeared on all forms

of the survey, stated simply Do you feel there is a difference

between altruism and heroism? followed by an open-ended item

asking for an elaboration. The overwhelming majority responded

that there was a difference (n 2,347 or 97.5%), while just 58

participants said they saw no difference between these two con-

7 Once the call for participation was posted on the technology blog,

response was overwhelming. Hundreds of responses were recorded in the

hours following the listing and a decision was made to leave the survey

open for several days to respect the efforts of the technology blog in

assisting the recruitment process. Rather than excluding some participants

(e.g., through random subsample, etc.), the approach described allows for

use of all available data while resolving the problem of a dramatically

overpowered sample.

Figure 1. Civil and martial heroism categories. N for items differs based on which form the item appeared in.

All within item 2 comparisons are significant at the .000 level in the direction of predominant bar.



11/16

cepts. The responses indicate that there is a significant perceived

difference between the ideas of heroism and altruism, 2(1)

2178.60, p .0001. Although a qualitative analysis of the open-

ended answers is outside the scope of this paper, representative

responses included sentiments such as, Altruism you are helpingothers out of morality and unselfishness. Heroism is the same

except in more extreme situations such as saving a life, Altruism

seems to involve helping a group of people directlythere is a

sacrifice, but it is more sharing what you have than sacrificing

yourself, and While heroic acts are always in some sense altru-

istic, an act of altruism isnt always heroic. Both the inferential

findings and descriptive information provide clear support for

Hypothesis 4, our assertion that there are fundamental perceived

differences between heroism and altruism.

Discussion

Study LimitationsReliability and response rates of Internet-based surveys have

been shown to be similar to that of more commonly used methods

of survey research (Mathy, Kerr, & Haydin, 2003). These methods

are able to reach more heterogeneous samples than is possible with

student subject pools, however, clear problems with sampling bias,

nonresponse rates, and generalizability remain (Birnbaum, 2004).

The present study benefits from having ample participation across

all age levels and robust responses from at least three continents.

However, the sample largely overrepresents the views of a White

American young adult males, limiting the generalizability of the

results to the broader population. Another critique offered by

participants was the American-centric viewpoint as the majority

heroic exemplars in the survey were drawn from American history

and were not as familiar to international participants. Future re-

search should focus on generating heroism instruments that can be

used effectively with diverse, international samples, and efforts toreplicate these findings across a range of other samples.

Further, some respondents noted in open-ended follow-up ques-

tions the desire to respond to each forced categorization item as

having various degrees of each attribute (i.e., using slider scales).

While the large sample in this study probably corrects for some of

the measurement error associated with this problem, future efforts

should examine if the proportions reported here are similar when

participants are allowed to freely assign values to all three attribute

categories.

Justification of Risk Constrains the Taxonomy

The support offered here for the insufficient justification argu-

ment provides insight across at least two levels. First, it demon-

strates that there are central criteria beyond risk type that are

important in the publics ascription of heroic statusas risk type

alone explained less than half of the variance in the model. Second,

it begins to meaningfully constrain the original taxonomy of 12

proposed heroic subtypes. While popular media often ascribe

heroic status to explorers and discoverers, the risks involved are

sufficiently justified. For example Amelia Earhart is frequently

cited in the news media and opinion polling as a heroic figure

(Williams, 1995), yet our findings suggest that the general public

does not view such acts as representative of the most central idea

of heroism. Even if these activities do not violate the operational

Figure 2. Social heroism categories. N for items differs based on which form the item appeared in. All within

item 2 comparisons are significant at the .000 level in the direction of predominant bar.



12/16

definition of heroism offered here in terms of extrinsic gain, the

level of intrinsic gain (satisfaction of achievement need, enjoyment

of arousal associated with risk, etc.) is high enough that the choiceto engage in these activities is seen as justified by the casual

observer. Broadly speaking, the descriptive and inferential results

suggest that the taxonomy can be reduced to six primary heroic

types: civil, martial, good Samaritans, politico-religious figures,

whistleblowers, and those who challenge problematic bureaucra-

cies. The social heroism categories that remain in this condensed

taxonomy have in common the perception of unjustified risk.

Directions for Future Research

The Role of Prototypicality, Velocity, and

Extension Neglect

We have argued that the less obvious form of heroismsocial

or principle driven heroisminvolves risks that should result in

the public rating actors in this category as being at least as heroic

as those in the physical risk arenas. However, the data presented

here and elsewhere (Rate et al., 2007) support the opposite con-

clusionthat the physical risk hero is viewed as more heroic than

social heroes. These findings suggest that the prototypicality or

centrality (Gardenfors, 2004; Rosch, 1975, 1978) of the immediate

physical risk associated with martial and civil heroism overrides

considerations more closely associated with social heroism such as

overall risk accepted, length of risk period, and so forth. Although

the connections to theories of prototype are clear, the underlying

Table 3

A Priori Heroic Items From Survey and a Posteriori Category Reordering

a Posteriori Categories Situation item rating Mean SD Narrative item rating Mean SD

Physical Peril, Not Duty BoundUnjustified risk 1. Civilian fire rescue 4.68 0.60 Robert Falconer 4.64 0.63

2. Wrestling gun from robber 4.43 0.85 Charles Carbonell 4.39 0.81

3. Criminal risks life 3.69 1.09 Joaquin Murietta 2.20 1.13

4. Bystander breaks up fight 3.10 0.97 Richard Crafton 4.20 0.795. Vigilante nabs criminals 2.44 1.17 Nat N. Kinney 2.36 1.01

Social Sacrifice, Defies SystemsUnjustified risk 1. Good Samaritan 4.401 0.83 Holocaust Rescuer2 4.071 0.96

2. Politico-religious figure 4.09 0.93 Martin L. King, Jr. 4.41 0.843. Bureaucracy hero 3.67 0.93 Dr. Edward Tolman 3.48 1.044. Whistleblower 3.31 1.14 Christina Maslach 3.06 1.125. Religious figure 3.27 1.22 The Dalai Lama 2.66 1.26

6. Martyr 2.31 1.28 Socrates 3.42 1.157. Political or military leader 3.08 1.22 Abraham Lincoln 3.03 1.25

Physical Peril, Duty BoundJustified risk 1. Soldier dies so others escape 4.65 0.78 CPL. Jason Dunham 4.63 0.75

2. Soldier rescues fallen buddy 4.16 0.96 SGT. Paul Smith 4.18 1.003. Police shot shielding partner 3.85 1.11 Dept. Alan Inzer 3.27 1.24

4. Soldier refuses info to enemy 3.54 1.11 Nathan Hale 3.49 1.185. Fire fighter rescues civilian 3.68 1.11 Richard Rescorla3 4.28 0.96

Social Sacrifice, Defies RealityJustified risk 1. Scientific discovery 3.20 1.14 Marie Curie 2.46 1.25

2. Odds beater/underdog 2.84 1.11 Lance Armstrong 2.48 1.153. Adventurer/explorer 2.85 1.09 Lewis & Clark 4 2.80 1.13

Note. Italicized items explore the relationship between criminality and heroism. These items were omitted from regression analysis.1 Both items recoded as physical risk, nonduty bound in later analyses after disagreement among authors and applying regression diagnostics. 2 GoodSamaritan wording variation: situation item referred to individual holocaust rescuer, narrative item to a family assisting Jews in war-time. 3 Item wordingvariation: Recorla not a firefighter, but he was duty bound to save individuals from fire. 4 Example of U.S. focus for many items limiting multicultural/multinational application.

Table 4

Multivariate Model of Heroism: Risk Type and Risk Justification

Model of risk type and risk justification(with Holocaust rescuers recoded to physical risk)

Variable Coefficient (SE) t p

Intercept 3.25 (0.13) 25.22 .0001Physical peril 1.06 (0.17) 6.29 .0001Justified risk 0.39 (0.17) 2.32 0.0266

R2 .55

Note. Although our goal was to test the theoretical model rather thandevelop an empirically driven model, two items involving Christian Ho-

locaust survivors had been contentious among us. Where the conceptshould fall in terms of principle driven social heroism versus immediatephysical risk dimension remained in question (see elementary model ofheroism; Zimbardo, 2007b, p. 480). Using DFFITS to explore for points ofinfluence in the model with the size adjusted cutoff recommended byBelsley, Kuh, & Welsch (2004) the generic situation Holocaust Rescueritem approximated this cut point. Both the generic situation item andnarrative item for this concept were recoded as physical risk items. Threeother items (generic situation bystander intervention, narrative politico-religious figure, and generic situation martyr) that approximated or ex-ceeded this cut point were left unchanged because there was no disagree-ment amongst the authors about the conceptual space these items occupied.The final model presented here modestly reduced the influence of justifi-able risk (though it was still a significant predictor), while increasing theexplained variance over the initial model by 9%.



13/16

mechanisms that make the classical example of a person running

into a burning building to save someone particularly representative

are worth additional exploration, and offer some potential alterna-

tive explanations.

For instance, in the burning building example the events

unfold rapidly, the problems are of vital importance to the potential

victims, and the ultimate resolution is unknown. These are the veryingredients of suspense (Vorderer, Wulff, & Friedrichsen, 1996;

Zillmann, 1991), they contain the cinematic elements that engen-

der strong emotional responses in viewers and are compelling in

part because they involve high velocity decisions and actions

(Alwitt, 2002; Carroll, 1996; Chatman, 1978; de Wied, 1994;

Nomikos, Opton, & Averill, 1968). Would we have viewed the

New York Subway hero Wesley Autrey (Buckley, 2007) as heroic

if the train was a minute or two further out from the station?

Probably not. In fact, a very similar incident occurred about 2

years later, in which Chad Johnson hoisted a man who had fallen

onto the tracks to get back up on the platform with some time to

spare, yet this event received much less fanfare (Wilson, 2009). To

be sure, the relative risk and the proximity of the risk event in time

are related. However, can we really say that racing to save some-one from an oncoming train is truly more heroic than the actions

of Dr. Martin Luther King, Jr., who received frequent death

threats, was nearly fatally stabbed early in his career as a civil

rights leader, and was ultimately assassinated in an effort to

achieve equality for an entire group of people? A subtle alternative

to risk-type as the primary driving force in prototypicality judg-

ments may involve the observers tendency to focus on the aver-

aged risk and the peak/end risks associated with civil and martial

heroism, while neglecting the duration, scope, and summative risks

associated with social heroism (Kahneman, 2002).

Is Heroism Always Positive?Positive psychology has advanced a largely virtuous, prosocial

view of heroism (Peterson & Seligman, 2004). The picture is

probably more nuanced. While most newsworthy heroes comport

with a fairly clean ideal of the heroic actor, the range of indi-

viduals who engage in heroic acts includes people who routinely

engaged in malicious behavior, but in this particularinstance did

something positive (e.g., Jabar Gibson; see Franco & Zimbardo,

20062007; Harlow, 2005).

Further, the possibility that heroic action is actually a symptom

of psychopathology or maladjustment should be seriously exam-

ined. For example, Pallone and Hennessey (1998) advanced the

notion of heroic rescue fantasy as a motivating factor, and

personality characteristics such as narcissism might also accountfor some proportion of civil heroism. Ludvik Vaculik pointed at

the negative aspects of heroic dissentparticularly when engaged

in without a comprehensive understanding of the situation, nor a

reasonable expectation that these actions will afford political

leverage:

It is one thing if they imprison someone who knows exactly what he

is doing and why, and quite another when a young, immature person

lands in jail, more or less by accident . . . A mass psychosis of heroism

is a fine thing, provided there are in the vicinity some sober minds

who have access to information and contacts and who know whats to

be done afterward. (1990, pp. 30 & 32).

Further negative views of heroism can be drawn from the

military context. At least since WWII, deteriorating mental health

in soldiers has been associated with a period of overconfidence

that follows maximum combat efficiency and precedes emotional

exhaustion (see Swank & Marchand, 1946, chart p. 238). The

combination of intense group cohesion, dissociation, distorted time

perspective, and rage toward the enemy may set the stage forheroic acts that dramatically exceed normal ability and that

would probably not be undertaken if the combatant was not al-

ready stretched to a psychological breaking point (Dinter, 1985;

Zimbardo & Boyd, 1999).

The Call to Heroism, Heroization, and

Retrospective Bias

While the personality-interactionist-situationist debate will not

be settled here, it remains important to stress the role of the societal

and social tableau upon which the heros story unfoldsif for no

other reason than to better understand the interactions that may

occur. Examining the American Civil Rights Movement, it is clearthat there were numerous heroesmost unrecognizedwho were

integral to the effort. From our vantage of history, Martin Luther

King, Jr., stands as the personified symbol of this heroic endeavor

and the sacrifices of many are epitomized in his individual death.

Yet, in the lead up to desegregation, any number of other figures

might have taken on this mantle (e.g., Ella Baker; Barnett, 1993).

In addition to the level of moral development and personality type,

heroism most probably involves a number of other personal ele-

mentsleadership style (Bennis, 2007), aptitude to address the

problem at hand (Glazer & Glazer, 1999) and whether or not the

person looks the part (Klapp, 1954). There are also external

elementsthe presence or absence of a situation that calls for

heroic action (Franco & Zimbardo, 20062007; Siklova, 2004;

Zimbardo, 2007b), the availability of resources, and capriciousnessof chance. For example, Klapp (1954) noted that:

Rationality, therefore should not be stressed as a factor in recognition

of heroes . . . most of such typing probably occurs by a spontaneous

popular definition in which there is little reflective thought . . . indeed,

rational procedures often only certify ex post facto a hero or anti hero

who has already been chosen by the public. Among the important

nonrational processes which help to form heroes and antiheroes are

gossip, rumor, propaganda, journalism, guilt by association, social

crisis mentality, and the accidents of publicity and opportunity which

have helped make some men famous and obscured others equally

deserving of credit or infamy (p. 59).

Further, in social heroism in particular there appears to be a

dynamic interaction between situations ostensibly calling for

heroism and the hero, resulting in a gradual progression toward the

acceptance of the heroic mantle (Howerth, 1935; Vaculik, 1990).

For example in discussing the life trajectory of whistleblowers,

Glazer and Glazer (1999) noted, Theirs were stories of women

pushing aside feelings of vulnerability to forge links and create

allies in the struggle to combat powerful adversaries . . . in turn,

Penny Newman and hundreds of others like her found their lives

transformed as they became recognized community leaders (p.

290).

The process of heroization points to one obvious problem

with retrospective studiesthese efforts capture information about



14/16

the hero at the end of a transformation toward intrepidityone

that may be powerful enough to affect personality. Further, utiliz-

ing major awards to retrospectively identify heroes (e.g., Walker,

Frimer, & Dunlop, 2010), rather than independently defining and

identifying heroic actors will inevitably fail to detect important

exemplars who may systematically differ from those obtaining

awards. This point cannot be underscored strongly enough. Forexample, Mohandas Ghandi, one of the most widely acclaimed

heroes of the 20th century, did not receive the Nobel Prize despite

being nominated five timesprobably because of cultural differ-

ences or political pressures felt by the Nobel committee (Tnnes-

son, 1999).

Resolution of Dissonance as the Power of

Heroic Action

While we have a historical appreciation of the controversy

surrounding heroic political and religious figures like Dr. Mar-

tin Luther King, Jr., Gandhi, and Jesus, we easily forget the

visceral reactions these individuals engendered in their time.

Even their supporters questioned their motives, while their

enemies loathed them (as witnessed by their untimely deaths).

In each case, it was not until some time had passed that the full

extent of their impact was understood, positively reappraised by

societyultimately leading to the label hero being posthu-

mously accepted widely (and their tumultuous journeys and

personal histories reified to comport with the label). The find-

ings of Monin, et al. (2008) suggest that morally rebellious

individuals are not successful in forcing involved actors to

consciously acknowledge that their own involvement/perfor-

mance may have been less than ethical. However, the true

power (and perhaps the final measure of success) of a social

hero is that their actions can ultimately guide us through the

dissonance, which they themselves produced, to embrace achallenging new set of values that has the potential to drive

further constructive action. It can be argued that at its most

extreme, it is the heros death or serious injury that forces her

opponents to reconsider the heros unpopular position and in-

sufficiently justified actionsleaving a soft-spot for this in-

dividual that allows for reevaluation of the issue in contention,

ultimately winning over some of her enemies to the justness of

the cause in question.

Fostering the Heroic Imagination

The construct of heroic imagination is central to our view of

heroism, although it remains largely theoretical and has not

been adequately characterized to date. Our initial work in this

area suggests that the heroic imagination functions in three

distinct ways: how heroes are imagined in classic writings and

by the general public; a mental state of anticipation and read-

iness for any person to act heroically when opportunities arise

calling for heroic actionsas a contrast with the hostile imag-

ination, or the psychology of enmity, which instills fear and

hatred of enemies; and, at least in some individuals, the ability

to envision and communicate a new way of ordering a social

system or an entire society.

Perhaps the most extreme views of heroism come from

classic writers, such as Thomas Carlyle, in his treatise on

Heroes and Hero Worship, written in 1840 (Carlyle, 1840). In

such conceptions, heroes are divine; they are the light that

enlightens the darkness of the world . . . as a natural luminary

shining by the gift of heaven. (p. 2). Carlyle goes on to

characterize how we should be imagining and worshiping he-

roes as Gods gift to humanity; Hero-worship, heartfelt pros-

trate admiration, submission, burning, boundless, for a noblestgodlike form of manis not that the germ of Christianity

itself? This rather extreme view of the nobility and divinity of

heroes is then extended to heroes as prophets, as poets, as

priests, as men of letters, and finally as kings. Joseph Campbell

(2003, 2008) modernized such views, but still put heroes on a

tall pedestal typically featuring male warriors, like Achilles,

and male adventurers, such as Odysseus, traveling on their

heroic paths. Even in modern times, there probably remains a

substantial divide between peoples assumptions about heroes

and what these heroes are like in real life. We have argued that

the perpetuation of the myth of the heroic elect does society

a disservice because it prevents the average citizen from

considering their own heroic potential.When reviewing the propaganda that most nations use to instill

hatred of their chosen enemies, Sam Keen (in Faces of the

Enemy, Keen, 1991) coined the term hostile imagination as the

desired outcome of these systemic attempts by nation states to

make their citizens hate and then want to kill anyone who fits the

characterization as enemy. Hostile imagination includes thinking

of other people as objects, as unworthy, as less than human, in

short, dehumanizing others. He argues that by instilling a psycho-

logical state of enmity in the general public they are more likely to

support war against the enemy selected by their leaders and to send

their sons into battle with that evil other. In this view, any of us can

be seduced into becoming perpetrators of evil by situational and

systemic forces acting upon us. The imperative becomes discov-

ering how to limit, constrain, and prevent those situational and

systemic forces that propel some of us toward social pathology,

and moreover, we have argued that societies should foster a

heroic imagination in their citizens as an antidote to evil (Zim-

bardo, 2007b; see also Franco & Zimbardo, 20062007; Zim-

bardo, 2006; Zimbardo, 2007a).

In this sense, the idea of heroic imagination, can be seen as

mind-set, a collection of attitudes about helping others in need,

beginning with caring for others in compassionate ways, but also

moving toward a willingness to sacrifice or take risks on behalf of

others or in defense of a moral cause. This conveys the message

that every person has the potential to act heroically. Anecdotal

evidence suggests that those who engage their heroic imagination

are making themselves aware of opportunities where they can helpothers in need, and may be more prepared to accept and transcend

the consequences associated with the heroic decision in complex

situations (Franco & Zimbardo, 20062007).

Finally, a third sense of the term heroic imagination stems

from the bold reinterpretations of societal order offered by some

social heroes. For example, Martin Luther King, Jr.s, statement, I

have dream is at its core a reimagining of life without the

constraints of apartheid. Moreover, he recognized that many of the

people listening to that speech had lost the ability imagine a better

world, or to believe that their own acceptance of risky actions had

the power to foster change.



15/16

Even in our moment in history, our heroes are not obvious, those

that we may later uphold as the paragons of human behavior are

currently subject to daily ebb and flow of support from their

adherents and distain from those in opposition.8 Our understanding

of historic figures who have attained this status has been shaped by

time, distance, and by the communal validation involved in the

social construction of their heroism. The World Wars of the lastcentury had a dramatic impact on our conceptualization of heroism

and as our society shifts increasingly toward a highly networked,

digitized future, the question of what the term hero will mean for

this generation is yet to be answered.

8 As this article was going to press, the issue of whether or not Mr. Julian

Assange, founder of the whistleblower Website Wikileaks, was being

widely discussed in the halls of government, in the press, and in social

media forums. Several discussions of his actions posed headline questions

like Is Assange a Hero or Villain? Some members of the U.S. Congress

have called for his arrest for publishing state secrets, while faculty of

journalism schools have called for him not to be prosecuted to preserve free

speech rights. While we do not offer an opinion on if Julian Assange is a

hero or not, it is an important and timely example of how contentious the

ascription of nonheroic or heroic status is in the moment.

References

Algoe, S. B., & Haidt, J. (2009). Witnessing excellence in action: The

other-praising emotions of elevation, gratitude, and admiration. Jour-

nal of Positive Psychology, 4, 105127.

Allison, S. T., & Goethals, G. R. (2010). Heroes: What they do and why we

need them. New York: Oxford University Press.

Alwitt, L. F. (2002). Suspense and advertising responses. Journal of

Consumer Psychology, 12, 3549.

Arendt, H. (1963/1994). Eichmann in Jerusalem: A report on the banality

of evil (Revised and enlarged edition). New York: Penguin.

Armed Forces Act, 10 U.S. C. 37413753 (2000).

Barnett, B. M. (1993). Invisible southern black women leaders in the civil

rights movement: The triple constraints of gender, race, and class.

Gender & Society, 7, 162182.

Becker, S. W., & Eagly, A. H. (2004). The heroism of women and men.

American Psychologist, 59, 163178.

Belsley, D. A., Kuh, E., & Welsch, R. E. (2004). Regression diagnostics:

Identifying Influential Data and Sources of Collinearity. New York:

Wiley.

Bennis, W. (2007). The challenges of leadership in the modern world:

Introduction to the special issue. American Psychologist, 62, 25.

Birnbaum, M. H. (2004). Human research and data collection via the

internet. Annual Review of Psychology, 55, 803832.

Braudy, L. (1997). The frenzy of renown: Fame and its history. New York:

Vintage Books.Buckley, C. (2007, January 3). Man is rescued by stranger on subway

tracks. The New York Times. Retrieved from http://www.nytimes.com/

2007/01/03/nyregion/03life.html

Campbell, J. (2003). The heros journey (3rd ed.). Novato, CA: New World

Library.

Campbell, J. (2008). The hero with a thousand faces (3rd ed.). Novato, CA:

New World Library.

Cardwell, D. (2007, January 5). Subway rescuer receives the citys highest award.

The New York Times. Retrieved from http://www.nytimes.com/2007/01/05/

nyregion/05life.html?_r2&scp5&sqwesley%20autrey&stcse

Carlyle, T. (1840). Heroes, hero workship: The heroic in history. New

York: A. L. Burt.

Carroll, N. (1996). The paradox of suspense. In P. Vroderer, H. J. Wulff,

& M. Friedrichsen (Eds.), Suspense: Conceptualizations, theoretical

analyses, and empirical explorations. Mahwah, NJ: Laur

heroism franco blau zimbardo

Documents